[GitHub] piyushghai commented on a change in pull request #11626: [MXNET-651] MXNet Model Backwards Compatibility Checker
piyushghai commented on a change in pull request #11626: [MXNET-651] MXNet Model Backwards Compatibility Checker URL: https://github.com/apache/incubator-mxnet/pull/11626#discussion_r206270486 ## File path: tests/nightly/model_backwards_compatibility_check/train_mxnet_legacy_models.sh ## @@ -57,10 +57,10 @@ do fi ## If MXNet major version starts with a number >=1. with a wildcard match for the minor version numbers - if [[ $version = [1-9]* ]] + if [[ $version = [1-9].[0-9].[0-9] ]] Review comment: Good point. Done 16d320a926d4f23265f730d2fc197ecec5bc4a2e. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sad- commented on issue #7652: Need a clear and thorough tutorial for writing custom operators in c++ with nnvm
sad- commented on issue #7652: Need a clear and thorough tutorial for writing custom operators in c++ with nnvm URL: https://github.com/apache/incubator-mxnet/issues/7652#issuecomment-408962967 @zheng-da can we close out this issue now? thanks This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] Roshrini commented on issue #11866: [MXNET-531] Add download util
Roshrini commented on issue #11866: [MXNET-531] Add download util URL: https://github.com/apache/incubator-mxnet/pull/11866#issuecomment-408965859 @yzhliu @nswamy Can you review this PR and approve it if ready? tracking this one for MXNet 1.3 release. Thanks! This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] Roshrini commented on issue #11753: [MXNET-531] RNN Example for Scala
Roshrini commented on issue #11753: [MXNET-531] RNN Example for Scala URL: https://github.com/apache/incubator-mxnet/pull/11753#issuecomment-408965920 @yzhliu @nswamy Can you review this PR and approve it if ready? tracking this one for MXNet 1.3 release. Thanks! This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] Roshrini commented on issue #11621: [MXNET-531] NeuralStyle Example for Scala
Roshrini commented on issue #11621: [MXNET-531] NeuralStyle Example for Scala URL: https://github.com/apache/incubator-mxnet/pull/11621#issuecomment-408965978 @yzhliu @nswamy Can you review this PR and approve it if ready? tracking this one for MXNet 1.3 release. Thanks! This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] Ishitori commented on issue #10219: Pytorch models inference by mxnet
Ishitori commented on issue #10219: Pytorch models inference by mxnet URL: https://github.com/apache/incubator-mxnet/issues/10219#issuecomment-408999064 @qingyuanxingsi, did you try the solution above? Did it help? Here is the tutorial on how to export PyTorch model to ONNX: https://github.com/onnx/tutorials/blob/master/tutorials/PytorchOnnxExport.ipynb Here is the tutorial on how to import ONNX model to MXNet (python, but C++ should be similar): https://github.com/onnx/tutorials/blob/master/tutorials/OnnxMxnetImport.ipynb This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] y12uc231 opened a new issue #11932: Not Found Links in the doc of MXNet : Training with multiple GPUs
y12uc231 opened a new issue #11932: Not Found Links in the doc of MXNet : Training with multiple GPUs URL: https://github.com/apache/incubator-mxnet/issues/11932 Doc Link : https://mxnet.incubator.apache.org/versions/master/faq/model_parallel_lstm.html There are some hyperlinks that dont work here :- https://github.com/apache/incubator-mxnet/blob/master/example/model-parallel/lstm/lstm.py https://github.com/apache/incubator-mxnet/blob/master/example/model-parallel/lstm/lstm.py#L171 https://github.com/apache/incubator-mxnet/blob/master/example/model-parallel/lstm/lstm_ptb.py#L97-L102 Will appreciate if you fix these. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #8583: Improve custom cross-entropy operator example
ThomasDelteil commented on issue #8583: Improve custom cross-entropy operator example URL: https://github.com/apache/incubator-mxnet/issues/8583#issuecomment-409013359 @eric-haibin-lin it looks like the example has been updated to use ndarray rather than numpy. Can you please close? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sad- commented on issue #9051: [Android, v0.11.0] The result of mxnet_predict is wrong! AND how to compile the amalgamation in Android Studio?
sad- commented on issue #9051: [Android,v0.11.0] The result of mxnet_predict is wrong! AND how to compile the amalgamation in Android Studio? URL: https://github.com/apache/incubator-mxnet/issues/9051#issuecomment-408971783 compiling amalgamation to work in android studio is addressed here: https://github.com/apache/incubator-mxnet/issues/2133 @nswamy can we close out this issue? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #11417: libomp.so dependency (need REAL fix)
ThomasDelteil commented on issue #11417: libomp.so dependency (need REAL fix) URL: https://github.com/apache/incubator-mxnet/issues/11417#issuecomment-408974322 @sandeep-krishnamurthy ping @lebeg any updated on whether openmp should or shouldn't be bundled with MXNet? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] szha commented on issue #11509: Hanging flaky test test_operator.test_norm @ Python 3: GPU Win
szha commented on issue #11509: Hanging flaky test test_operator.test_norm @ Python 3: GPU Win URL: https://github.com/apache/incubator-mxnet/issues/11509#issuecomment-408988191 Hao and I weren't able to reproduce the error on the (supposedly) same instance type that CI slaves use, using the same docker building logic. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] marcoabreu commented on issue #11509: Hanging flaky test test_operator.test_norm @ Python 3: GPU Win
marcoabreu commented on issue #11509: Hanging flaky test test_operator.test_norm @ Python 3: GPU Win URL: https://github.com/apache/incubator-mxnet/issues/11509#issuecomment-408988244 G3 Are you running with our docker setup? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] szha commented on issue #11509: Hanging flaky test test_operator.test_norm @ Python 3: GPU Win
szha commented on issue #11509: Hanging flaky test test_operator.test_norm @ Python 3: GPU Win URL: https://github.com/apache/incubator-mxnet/issues/11509#issuecomment-408996025 @marcoabreu Hao and I have been using my branch and I imagine @leezu is doing the same. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] marcoabreu commented on a change in pull request #11626: [MXNET-651] MXNet Model Backwards Compatibility Checker
marcoabreu commented on a change in pull request #11626: [MXNET-651] MXNet
Model Backwards Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#discussion_r206315214
##
File path:
tests/nightly/model_backwards_compatibility_check/train_mxnet_legacy_models.sh
##
@@ -0,0 +1,69 @@
+#!/usr/bin/env bash
+
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+#Author: Piyush Ghai
+
+run_models() {
+ echo '=='
+ echo "Running training files and preparing models"
+ echo '=='
+ python model_backwards_compat_train.py
+ echo '=='
+}
+
+install_mxnet() {
+ version=$1
+ echo "Installing MXNet "$version
+ pip install mxnet==$version
+}
+
+install_boto3(){
+ echo "Intalling boto3"
+ pip install boto3
+}
+
+echo `pwd`
+cd tests/nightly/model_backwards_compatibility_check
+echo `pwd`
+
+install_boto3
+
+## Fetch the latest release tags, filtering out 'rcs' and filtering out some
other irrelevant ones
+## This list is sorted in descending order chronologically. Keeping n = 5 for
a precautionary check.
+## Sample output for the below git tag command is : 1.2.0 utils 1.1.0 1.0.0
0.12.1
+previous_versions=($(git tag --sort=-creatordate | grep --invert-match rc |
head -n 5))
+count=0
+for version in ${previous_versions[*]}
+do
+ # We just need to train the previous two versions. This logic can be
changed later on as welll.
+ if [[ "$count" -gt 1 ]]
Review comment:
As stated in-person, we have to train everything under the same major
version.
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] aaronmarkham commented on issue #1722: problem ：compile mxnet with hdfs
aaronmarkham commented on issue #1722: problem ：compile mxnet with hdfs URL: https://github.com/apache/incubator-mxnet/issues/1722#issuecomment-409008895 A new FAQ page would work… and then i'd mention it in the build from source sections on pages in the /install directory too, linking back to that FAQ. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #7717: Subpixel convolution(state of art) implementation rather than using Deconvolution.
ThomasDelteil commented on issue #7717: Subpixel convolution(state of art) implementation rather than using Deconvolution. URL: https://github.com/apache/incubator-mxnet/issues/7717#issuecomment-409008465 @chowkamlee81 are you still working on that? If not please consider closing. Can you provide your a full reproducible code sample and I'll try to investigate why your loss is not changing. One possibility is that you haven't set your learning rate properly, or your gradients are not flowing back from your loss. You can have a look at the GluonCV FCN implementation as a reference: https://gluon-cv.mxnet.io/build/examples_segmentation/train_fcn.html#sphx-glr-build-examples-segmentation-train-fcn-py @sandeep-krishnamurthy can you add "Pending Requester Info" tag? And remove the Example tag This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #8729: Build amalgamation using a docker
ThomasDelteil commented on issue #8729: Build amalgamation using a docker URL: https://github.com/apache/incubator-mxnet/issues/8729#issuecomment-408957075 @OferRosenberg please consider closing if your question has been answered. Summary of the conversation: - There is no amalgation build for Android - Full MXNet cross-compilation is possible, check out the build scripts/docker images in in ci/build.py - CMake needs to be improved, this is a WIP, addressed in #11148 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #9915: Our FCN-xs result on Pascal VOC Validation
ThomasDelteil commented on issue #9915: Our FCN-xs result on Pascal VOC Validation URL: https://github.com/apache/incubator-mxnet/issues/9915#issuecomment-408960146 The accuracy of the FCN models offered on Gluon-CV are reported here: https://gluon-cv.mxnet.io/model_zoo/index.html#semantic-segmentation Name | Method | pixAcc | mIoU | Command | log -- | -- | -- | -- | -- | -- fcn_resnet50_voc | FCN [6] | N/A | 69.4 | shell script | log fcn_resnet101_voc | FCN [6] | N/A | 70.9 | shell script | log @indhub can we please close this question? Thanks! This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] taliesinb commented on issue #11928: Generalized reshape_like operator
taliesinb commented on issue #11928: Generalized reshape_like operator URL: https://github.com/apache/incubator-mxnet/pull/11928#issuecomment-408963960 Please see this 'sister' issue that designs what is effectively the dual of this feature: https://discuss.mxnet.io/t/proposal-new-merge-dims-reshaping-op/1524 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sad- commented on issue #8474: How to train variable input shape CNN?
sad- commented on issue #8474: How to train variable input shape CNN? URL: https://github.com/apache/incubator-mxnet/issues/8474#issuecomment-408963756 @ijkguo could we close out this issue? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] safrooze commented on issue #11150: Hope for adding sample about mx.rnn.ConvGRUCell()
safrooze commented on issue #11150: Hope for adding sample about mx.rnn.ConvGRUCell() URL: https://github.com/apache/incubator-mxnet/issues/11150#issuecomment-408965653 @Feywell Thanks for posting your question on discuss forum. Please close this issue I'll help you on the forum to resolve your question. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] aaronmarkham commented on issue #11049: Add linux and macos MKLDNN Building Instruction
aaronmarkham commented on issue #11049: Add linux and macos MKLDNN Building Instruction URL: https://github.com/apache/incubator-mxnet/pull/11049#issuecomment-408969056 I have it here: https://issues.apache.org/jira/browse/MXNET-746 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] leezu commented on issue #11509: Hanging flaky test test_operator.test_norm @ Python 3: GPU Win
leezu commented on issue #11509: Hanging flaky test test_operator.test_norm @ Python 3: GPU Win URL: https://github.com/apache/incubator-mxnet/issues/11509#issuecomment-408985161 I can't reproduce this on p3 and p2 instances so far. I keep running this test in a loop for now. Do the tests run on a different instance type? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #1722: problem ：compile mxnet with hdfs
ThomasDelteil commented on issue #1722: problem ：compile mxnet with hdfs URL: https://github.com/apache/incubator-mxnet/issues/1722#issuecomment-409002324 @aaronmarkham where would be the best place to add some documentation on how to build with HDFS support? Currently it seems only documented in the cinnebts of some Makefile. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #8156: Why mx-rcnn doesn't has proposal_target (cuda or C++) layer?
ThomasDelteil commented on issue #8156: Why mx-rcnn doesn't has proposal_target (cuda or C++) layer? URL: https://github.com/apache/incubator-mxnet/issues/8156#issuecomment-409012392 You can find a hybridizable Gluon implementation of Faster RCNN here: https://gluon-cv.mxnet.io/build/examples_detection/train_faster_rcnn_voc.html#sphx-glr-build-examples-detection-train-faster-rcnn-voc-py From a quick look, it seems that the layer you are looking for is defined here: https://github.com/dmlc/gluon-cv/blob/master/gluoncv/model_zoo/faster_rcnn/rcnn_target.py @indhub can you please close the issue? @coolbrain @bharatsingh430 if that doesn't answer your question, please open a new thread on https://discuss.mxnet.io backlinking to this issue. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #9217: Installing GPU support on Mac
ThomasDelteil commented on issue #9217: Installing GPU support on Mac URL: https://github.com/apache/incubator-mxnet/issues/9217#issuecomment-409024466 Installation steps provided from this discussion have been made available on the mxnet website there: http://mxnet.incubator.apache.org/install/index.html?platform=MacOS=Python=GPU @szha could you please close this issue? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sandeep-krishnamurthy closed issue #9018: Seg Fault on MNIST example - C++
sandeep-krishnamurthy closed issue #9018: Seg Fault on MNIST example - C++ URL: https://github.com/apache/incubator-mxnet/issues/9018 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] zhreshold closed issue #6545: SSD support for arbitrary shapes at prediction (without resize)
zhreshold closed issue #6545: SSD support for arbitrary shapes at prediction (without resize) URL: https://github.com/apache/incubator-mxnet/issues/6545 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] zhreshold commented on issue #6545: SSD support for arbitrary shapes at prediction (without resize)
zhreshold commented on issue #6545: SSD support for arbitrary shapes at prediction (without resize) URL: https://github.com/apache/incubator-mxnet/issues/6545#issuecomment-409040001 Checkout our new gluon-cv examples: https://gluon-cv.mxnet.io/build/examples_detection/demo_ssd.html#sphx-glr-build-examples-detection-demo-ssd-py which fully support arbitrary shape images. I am closing this now. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] Roshrini commented on issue #11933: [Automated Publish] Help needed to run restricted CI
Roshrini commented on issue #11933: [Automated Publish] Help needed to run restricted CI URL: https://github.com/apache/incubator-mxnet/issues/11933#issuecomment-409041282 @sandeep-krishnamurthy Could you please label: CI, Scala This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sad- commented on issue #7973: C++ demo memory release problem
sad- commented on issue #7973: C++ demo memory release problem URL: https://github.com/apache/incubator-mxnet/issues/7973#issuecomment-409047775 @junxigauss is this still an issue on mxnet v1.2? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] pengzhao-intel commented on issue #11049: Add linux and macos MKLDNN Building Instruction
pengzhao-intel commented on issue #11049: Add linux and macos MKLDNN Building Instruction URL: https://github.com/apache/incubator-mxnet/pull/11049#issuecomment-409057867 @szha @aaronmarkham really thanks for your help. We will track the issue and update the doc. Feel free to ping us if anything needs to cover. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] szha closed issue #9588: metric should have TP FP TN FN Precision Recall F1 for both macro and micro versions
szha closed issue #9588: metric should have TP FP TN FN Precision Recall F1 for both macro and micro versions URL: https://github.com/apache/incubator-mxnet/issues/9588 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] szha commented on issue #9588: metric should have TP FP TN FN Precision Recall F1 for both macro and micro versions
szha commented on issue #9588: metric should have TP FP TN FN Precision Recall F1 for both macro and micro versions URL: https://github.com/apache/incubator-mxnet/issues/9588#issuecomment-409060120 Indeed This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] elofgren commented on issue #11890: Error During Make - undefined reference
elofgren commented on issue #11890: Error During Make - undefined reference URL: https://github.com/apache/incubator-mxnet/issues/11890#issuecomment-409018204 Just attempted to reinstall OpenCV following this guide: https://gist.github.com/dnordgren/665af430064cdbaa9a31 The error that followed: > tools/im2rec.cc: In function ‘int main(int, char**)’: tools/im2rec.cc:88:20: error: ‘CV_LOAD_IMAGE_COLOR’ was not declared in this scope int color_mode = CV_LOAD_IMAGE_COLOR; ^ tools/im2rec.cc:90:22: error: ‘CV_INTER_LINEAR’ was not declared in this scope int inter_method = CV_INTER_LINEAR; ^ tools/im2rec.cc:195:31: error: ‘CV_IMWRITE_PNG_COMPRESSION’ was not declared in this scope encode_params.push_back(CV_IMWRITE_PNG_COMPRESSION); ^ tools/im2rec.cc:199:31: error: ‘CV_IMWRITE_JPEG_QUALITY’ was not declared in this scope encode_params.push_back(CV_IMWRITE_JPEG_QUALITY); ^ make: *** [bin/im2rec] Error 1 This seems like it's pretty clearly an OpenCV problem? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #9018: Seg Fault on MNIST example - C++
ThomasDelteil commented on issue #9018: Seg Fault on MNIST example - C++ URL: https://github.com/apache/incubator-mxnet/issues/9018#issuecomment-409022487 @Austin-Bolesta the CPP examples were in bad shape and have all been fixed and are now continuously tested. ``` ~/incubator-mxnet$ make -j $(nproc) USE_OPENCV=1 USE_BLAS=openblas USE_CPP_PACKAGE=1 ~/incubator-mxnet$ cd cpp-package/example ~/incubator-mxnet/cpp-package/example$ make all ~/incubator-mxnet/cpp-package/example$ cp ../../lib/libmxnet.so . ~/incubator-mxnet/cpp-package/example$ ./mlp_cpu [21:45:19] src/io/iter_mnist.cc:110: MNISTIter: load 6 images, shuffle=1, shape=(100,784) [21:45:19] src/io/iter_mnist.cc:110: MNISTIter: load 1 images, shuffle=1, shape=(100,784) [21:45:20] mlp_cpu.cpp:134: Epoch: 0 121704 samples/sec Accuracy: 0.1135 [21:45:21] mlp_cpu.cpp:134: Epoch: 1 125000 samples/sec Accuracy: 0.536 [21:45:21] mlp_cpu.cpp:134: Epoch: 2 133929 samples/sec Accuracy: 0.8278 [21:45:22] mlp_cpu.cpp:134: Epoch: 3 134228 samples/sec Accuracy: 0.8726 [21:45:22] mlp_cpu.cpp:134: Epoch: 4 131579 samples/sec Accuracy: 0.9041 [21:45:23] mlp_cpu.cpp:134: Epoch: 5 131291 samples/sec Accuracy: 0.9163 [21:45:23] mlp_cpu.cpp:134: Epoch: 6 131291 samples/sec Accuracy: 0.9226 [21:45:24] mlp_cpu.cpp:134: Epoch: 7 131004 samples/sec Accuracy: 0.9291 [21:45:24] mlp_cpu.cpp:134: Epoch: 8 131004 samples/sec Accuracy: 0.9334 [21:45:25] mlp_cpu.cpp:134: Epoch: 9 131579 samples/sec Accuracy: 0.937 ``` @marcoabreu can you please close? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil edited a comment on issue #9018: Seg Fault on MNIST example - C++
ThomasDelteil edited a comment on issue #9018: Seg Fault on MNIST example - C++ URL: https://github.com/apache/incubator-mxnet/issues/9018#issuecomment-409022487 @Austin-Bolesta the CPP examples were in bad shape and have all been fixed and are now continuously tested. ``` ~/incubator-mxnet$ make -j $(nproc) USE_OPENCV=1 USE_BLAS=openblas USE_CPP_PACKAGE=1 ~/incubator-mxnet$ cd cpp-package/example ~/incubator-mxnet/cpp-package/example$ make all ~/incubator-mxnet/cpp-package/example$ cp ../../lib/libmxnet.so . ~/incubator-mxnet/cpp-package/example$ ./mlp_cpu [21:45:19] src/io/iter_mnist.cc:110: MNISTIter: load 6 images, shuffle=1, shape=(100,784) [21:45:19] src/io/iter_mnist.cc:110: MNISTIter: load 1 images, shuffle=1, shape=(100,784) [21:45:20] mlp_cpu.cpp:134: Epoch: 0 121704 samples/sec Accuracy: 0.1135 [21:45:21] mlp_cpu.cpp:134: Epoch: 1 125000 samples/sec Accuracy: 0.536 [21:45:21] mlp_cpu.cpp:134: Epoch: 2 133929 samples/sec Accuracy: 0.8278 [21:45:22] mlp_cpu.cpp:134: Epoch: 3 134228 samples/sec Accuracy: 0.8726 [21:45:22] mlp_cpu.cpp:134: Epoch: 4 131579 samples/sec Accuracy: 0.9041 [21:45:23] mlp_cpu.cpp:134: Epoch: 5 131291 samples/sec Accuracy: 0.9163 [21:45:23] mlp_cpu.cpp:134: Epoch: 6 131291 samples/sec Accuracy: 0.9226 [21:45:24] mlp_cpu.cpp:134: Epoch: 7 131004 samples/sec Accuracy: 0.9291 [21:45:24] mlp_cpu.cpp:134: Epoch: 8 131004 samples/sec Accuracy: 0.9334 [21:45:25] mlp_cpu.cpp:134: Epoch: 9 131579 samples/sec Accuracy: 0.937 ``` @marcoabreu @indhub @sandeep-krishnamurthy can you please close? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil edited a comment on issue #10897: performance and error issue with INT8 quantization example
ThomasDelteil edited a comment on issue #10897: performance and error issue with INT8 quantization example URL: https://github.com/apache/incubator-mxnet/issues/10897#issuecomment-409028803 @reminisce have the docs been updated to let users know about the slowness and maybe explain the source of it? @reminisce can you remove the Example and Question tags now that the question has been answered Thanks! This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #10897: performance and error issue with INT8 quantization example
ThomasDelteil commented on issue #10897: performance and error issue with INT8 quantization example URL: https://github.com/apache/incubator-mxnet/issues/10897#issuecomment-409028803 @reminisce have the docs been updated to let users know about the slowness and maybe explain the source of it? Thanks! This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet] branch master updated: Improving documentation and error messages for Async distributed training with Gluon (#11910)
This is an automated email from the ASF dual-hosted git repository.
skm pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git
The following commit(s) were added to refs/heads/master by this push:
new 478b4a1 Improving documentation and error messages for Async
distributed training with Gluon (#11910)
478b4a1 is described below
commit 478b4a1a6f4ea93b291dd5ace3dc85e21d823453
Author: Rahul Huilgol
AuthorDate: Mon Jul 30 15:30:51 2018 -0700
Improving documentation and error messages for Async distributed training
with Gluon (#11910)
* Add description about update on kvstore
* add async check for gluon
* only raise error if user set update_on_kvstore
* fix condition
* add async nightly test
* fix case when no kvstore
* add example for trainer creation in doc
---
docs/faq/distributed_training.md| 21 ++--
python/mxnet/gluon/trainer.py | 8 ++-
tests/nightly/dist_async_kvstore.py | 48 +
3 files changed, 74 insertions(+), 3 deletions(-)
diff --git a/docs/faq/distributed_training.md b/docs/faq/distributed_training.md
index 70078ba..d4fa72d 100644
--- a/docs/faq/distributed_training.md
+++ b/docs/faq/distributed_training.md
@@ -73,6 +73,23 @@ These can be passed as arguments to the iterator.
You can look at
[example/gluon/image_classification.py](https://github.com/apache/incubator-mxnet/blob/master/example/gluon/image_classification.py)
to see an example usage.
+### Updating weights
+KVStore server supports two modes, one which aggregates the gradients and
updates the weights using those gradients, and second where the server only
aggregates gradients. In the latter case, when a worker process pulls from
kvstore, it gets the aggregated gradients. The worker then uses these gradients
and applies the weights locally.
+
+When using Gluon there is an option to choose between these modes by passing
`update_on_kvstore` variable when you create the
[Trainer](https://mxnet.incubator.apache.org/versions/master/api/python/gluon/gluon.html#mxnet.gluon.Trainer)
object like this:
+
+```
+trainer = gluon.Trainer(net.collect_params(), optimizer='sgd',
+optimizer_params={'learning_rate': opt.lr,
+ 'wd': opt.wd,
+ 'momentum': opt.momentum,
+ 'multi_precision': True},
+kvstore=kv,
+update_on_kvstore=True)
+```
+
+When using the symbolic interface, it performs the weight updates on the
server without the user having to do anything special.
+
### Different Modes of Distributed Training
Distributed training itself is enabled when kvstore creation string contains
the word `dist`.
@@ -86,9 +103,9 @@ In this mode, if a worker crashes, then it halts the
progress of all workers.
- `dist_async`: In asynchronous distributed training, the server receives
gradients from one worker and immediately updates its store, which it uses to
respond to any future pulls.
This means that a worker who finishes processing a batch can pull the current
parameters from server and start the next batch,
even if other workers haven't finished processing the earlier batch.
-This is faster than `dist_sync` but can take more epochs to converge.
-In `async` mode, it is required to pass an optimizer because in the absence of
an optimizer kvstore would replace the stored weights with received weights and
this doesn't make sense for training in asynchronous mode.
+This is faster than `dist_sync` because there is no cost of synchronization,
but can take more epochs to converge.
The update of weights is atomic, meaning no two updates happen on the same
weight at the same time. However, the order of updates is not guaranteed.
+In `async` mode, it is required to pass an optimizer because in the absence of
an optimizer kvstore would replace the stored weights with received weights and
this doesn't make sense for training in asynchronous mode. Hence, when using
Gluon with `async` mode we need to set `update_on_kvstore` to `True`.
- `dist_sync_device`: Same as `dist_sync` except that when there are multiple
GPUs being used on each node,
this mode aggregates gradients and updates weights on GPU while dist_sync does
so on CPU memory.
diff --git a/python/mxnet/gluon/trainer.py b/python/mxnet/gluon/trainer.py
index b426341..98a6878 100644
--- a/python/mxnet/gluon/trainer.py
+++ b/python/mxnet/gluon/trainer.py
@@ -187,6 +187,11 @@ class Trainer(object):
arg_arrays = {param.name: param.data(self._contexts[0]) for param
in self._params}
kvstore, update_on_kvstore = _create_kvstore(config['kvstore'],
len(self._contexts),
arg_arrays)
+if kvstore and 'async' in
[GitHub] anirudhacharya commented on issue #9136: Autograd in R-package
anirudhacharya commented on issue #9136: Autograd in R-package URL: https://github.com/apache/incubator-mxnet/issues/9136#issuecomment-409033572 @hetong007 there is no roadmap as of now, but yes please do remove the 'Question' label. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sandeep-krishnamurthy commented on issue #10955: [MXNET-422] Distributed training tutorial
sandeep-krishnamurthy commented on issue #10955: [MXNET-422] Distributed training tutorial URL: https://github.com/apache/incubator-mxnet/pull/10955#issuecomment-409033954 @indhub - Can you update the PR with rebasing and we can get it merged? This tutorial is useful for the users. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sandeep-krishnamurthy closed issue #8636: What differences between ImageRecordIter and ImageRecordIter_v1?
sandeep-krishnamurthy closed issue #8636: What differences between ImageRecordIter and ImageRecordIter_v1? URL: https://github.com/apache/incubator-mxnet/issues/8636 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] marcoabreu commented on a change in pull request #11626: [MXNET-651] MXNet Model Backwards Compatibility Checker
marcoabreu commented on a change in pull request #11626: [MXNET-651] MXNet
Model Backwards Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#discussion_r206345666
##
File path:
tests/nightly/model_backwards_compatibility_check/train_mxnet_legacy_models.sh
##
@@ -0,0 +1,82 @@
+#!/usr/bin/env bash
+
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+#Author: Piyush Ghai
+
+set -ex
+
+run_models() {
+ echo '=='
+ echo "Running training files and preparing models"
+ echo '=='
+ python model_backwards_compat_train.py
+ echo '=='
+}
+
+install_mxnet() {
+ version=$1
+ echo "Installing MXNet "$version
+ pip install mxnet==$version --user
+}
+
+## Cuts the string and gives only the major version part.
+## eg : 12.3.0 ---> 12
+get_major_version() {
+major=$(echo $1 | cut -d. -f1)
+echo $major
+}
+
+## We read the current major version from libinfo.py file. And we extract the
major version from it.
+curr_mxnet_version=$(grep -w "__version__" python/mxnet/libinfo.py | grep -o
'".*"' | sed 's/"//g')
+curr_major_version=$(get_major_version $curr_mxnet_version)
Review comment:
Please validate the output to ensure we get a notification/error if somebody
breaks this by accident.
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] vishaalkapoor opened a new pull request #11934: [MXAPPS-581] Straight Dope nightly fixes.
vishaalkapoor opened a new pull request #11934: [MXAPPS-581] Straight Dope nightly fixes. URL: https://github.com/apache/incubator-mxnet/pull/11934 * Enable 3 notebooks that were failing tests after making updates to the Straight Dope book. * Add pandas dependency required by one of these notebooks. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] jimdunn commented on issue #11808: remove mod from arity 2 version of load-checkpoint in clojure-package
jimdunn commented on issue #11808: remove mod from arity 2 version of load-checkpoint in clojure-package URL: https://github.com/apache/incubator-mxnet/pull/11808#issuecomment-409037602 Oh sure no problem. I'll take care of that in the next day or two. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sad- commented on issue #6545: SSD support for arbitrary shapes at prediction (without resize)
sad- commented on issue #6545: SSD support for arbitrary shapes at prediction (without resize) URL: https://github.com/apache/incubator-mxnet/issues/6545#issuecomment-409039583 @sandeep-krishnamurthy can we close out this issue actually? looks to have been addressed. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] safrooze edited a comment on issue #11913: Unexpectedly poor copy() performance
safrooze edited a comment on issue #11913: Unexpectedly poor copy() performance URL: https://github.com/apache/incubator-mxnet/issues/11913#issuecomment-409040744 @rahul003 Thanks for the comment. The results are identical, with or without the `nd.waitall()` because creating creating an empty array is very fast compared to the rest of the script. Nonetheless, I did update the description to include an `nd.waitall()` after array creation. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] safrooze edited a comment on issue #11913: Unexpectedly poor copy() performance
safrooze edited a comment on issue #11913: Unexpectedly poor copy() performance URL: https://github.com/apache/incubator-mxnet/issues/11913#issuecomment-409040744 @rahul003 Thanks for the comment. The results are identical, with or without the 1nd.waitall()` because creating creating an empty array is very fast compared to the rest of the network. Nonetheless, I did update the description to include an `nd.waitall()` after array creation. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] safrooze edited a comment on issue #11913: Unexpectedly poor copy() performance
safrooze edited a comment on issue #11913: Unexpectedly poor copy() performance URL: https://github.com/apache/incubator-mxnet/issues/11913#issuecomment-409040744 @rahul003 Thanks for the comment. The results are identical, with or without the `nd.waitall()` because creating creating an empty array is very fast compared to the rest of the network. Nonetheless, I did update the description to include an `nd.waitall()` after array creation. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sad- commented on issue #10049: Building MxNet on Win10 with CPP package, CUDA9 and MKL 2018
sad- commented on issue #10049: Building MxNet on Win10 with CPP package, CUDA9 and MKL 2018 URL: https://github.com/apache/incubator-mxnet/issues/10049#issuecomment-409040711 @sandeep-krishnamurthy could we close this issue. user seems to have resolved it. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] safrooze commented on issue #11913: Unexpectedly poor copy() performance
safrooze commented on issue #11913: Unexpectedly poor copy() performance URL: https://github.com/apache/incubator-mxnet/issues/11913#issuecomment-409040744 @rahul003 Thanks for the comment. The results are identical, with or without the nd.waitall() because creating creating an empty array is very fast compared to the rest of the network. Nonetheless, I did update the description to include an `nd.waitall()` after array creation. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sad- commented on issue #8959: Loss becomes nan when using correlation loss with MakeLoss on 2 images.
sad- commented on issue #8959: Loss becomes nan when using correlation loss with MakeLoss on 2 images. URL: https://github.com/apache/incubator-mxnet/issues/8959#issuecomment-409044211 @CorneliusHagmeister could you close this issue, if your issue is resolved? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] aaronmarkham commented on issue #11438: docs using xml in code blocks will improperly render
aaronmarkham commented on issue #11438: docs using xml in code blocks will improperly render URL: https://github.com/apache/incubator-mxnet/issues/11438#issuecomment-409045487 @andrewfayres and @lanking520 - They said they might be able to help figure out what is going on with it. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] CorneliusHagmeister closed issue #8959: Loss becomes nan when using correlation loss with MakeLoss on 2 images.
CorneliusHagmeister closed issue #8959: Loss becomes nan when using correlation loss with MakeLoss on 2 images. URL: https://github.com/apache/incubator-mxnet/issues/8959 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sandeep-krishnamurthy closed issue #8985: Torch Example Requires mx.th.randn and mx.symbol.TorchModule API
sandeep-krishnamurthy closed issue #8985: Torch Example Requires mx.th.randn and mx.symbol.TorchModule API URL: https://github.com/apache/incubator-mxnet/issues/8985 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] xinyu-intel opened a new pull request #11939: A hot fix for MKLDNN link in README
xinyu-intel opened a new pull request #11939: A hot fix for MKLDNN link in README URL: https://github.com/apache/incubator-mxnet/pull/11939 ## Description ## This is a hot fix for MKLDNN link in README.md. @pengzhao-intel ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [ ] The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant [JIRA issue](https://issues.apache.org/jira/projects/MXNET/issues) created (except PRs with tiny changes) - [x] Changes are complete (i.e. I finished coding on this PR) - [x] All changes have test coverage: - Unit tests are added for small changes to verify correctness (e.g. adding a new operator) - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore) - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL) - [x] Code is well-documented: - For user-facing API changes, API doc string has been updated. - For new C++ functions in header files, their functionalities and arguments are documented. - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable - Check the API doc at http://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html - [x] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] lanking520 commented on issue #2268: MXNet on Spark Roadmap
lanking520 commented on issue #2268: MXNet on Spark Roadmap URL: https://github.com/apache/incubator-mxnet/issues/2268#issuecomment-409017016 I have updated with a CWiki page here: https://cwiki.apache.org/confluence/display/MXNET/Scala+Project+Status Let's keep all TODOs into one place and also pick up from there. @yzhliu @nswamy let's close this for now. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] lanking520 commented on issue #4800: Proposed improvements for the mxnet Scala package
lanking520 commented on issue #4800: Proposed improvements for the mxnet Scala package URL: https://github.com/apache/incubator-mxnet/issues/4800#issuecomment-409017083 I have updated with a CWiki page here: https://cwiki.apache.org/confluence/display/MXNET/Scala+Project+Status Let's keep all TODOs into one place and also pick up from there. @yzhliu @nswamy let's close this for now. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] lanking520 opened a new pull request #11844: [MXNET-689] add DataDesc type for the Scala Package
lanking520 opened a new pull request #11844: [MXNET-689] add DataDesc type for the Scala Package URL: https://github.com/apache/incubator-mxnet/pull/11844 ## Description ## This PR includes massive changes on all Iterators that extends `DataIter`, two new fields are added (`dtype` and `layout`). @nswamy @yzhliu @andrewfayres It intends to solve some issues about the layout information missing in DataDesc. However I found it hard to have a general solution that both address backward compatibility as well as adding this information. Currently, there are several ways to address this issue: 1. Assume Shape == 2 follows `NT` format, Shape == 3 follows `TNC`, Shape == 4 follows `NCHW` and only change the implicit function in [here](https://github.com/apache/incubator-mxnet/blob/master/scala-package/core/src/main/scala/org/apache/mxnet/IO.scala#L358-L363). 2. This PR: add `Layout` and `DType` defs for all iterator, then users can pass these information to create `DataDesc` instead of `ListMap[Name, Shape]`. However, it breaks the input args for `NDArrayIter` as well as other iterators. Users are required to add more inputs. However, if the iterator are passed in with dynamic shapes (shapes are not determined), it will break the DataDesc checking scheme. Historical issues: - https://github.com/apache/incubator-mxnet/issues/11775 - [MLP problem](https://github.com/apache/incubator-mxnet/issues/10753) - [RNN problem](https://github.com/apache/incubator-mxnet/issues/11571) ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [x] The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant [JIRA issue](https://issues.apache.org/jira/projects/MXNET/issues) created (except PRs with tiny changes) - [x] Changes are complete (i.e. I finished coding on this PR) - [x] All changes have test coverage: - Unit tests are added for small changes to verify correctness (e.g. adding a new operator) - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore) - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL) - [ ] Code is well-documented: - For user-facing API changes, API doc string has been updated. - For new C++ functions in header files, their functionalities and arguments are documented. - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable - Check the API doc at http://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html - [x] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #11438: docs using xml in code blocks will improperly render
ThomasDelteil commented on issue #11438: docs using xml in code blocks will improperly render URL: https://github.com/apache/incubator-mxnet/issues/11438#issuecomment-409029184 @aaronmarkham is there further action required on this issue? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #9589: mx.metric should support top-K version of TP FP TN FN Precision Recall F1
ThomasDelteil commented on issue #9589: mx.metric should support top-K version of TP FP TN FN Precision Recall F1 URL: https://github.com/apache/incubator-mxnet/issues/9589#issuecomment-409030763 @szha Can we remove the Application tag, add the Feature Request tag instead? Also not sure why the Autograd tag is here, are there particular concern regarding autograd with this change? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil edited a comment on issue #9589: mx.metric should support top-K version of TP FP TN FN Precision Recall F1
ThomasDelteil edited a comment on issue #9589: mx.metric should support top-K version of TP FP TN FN Precision Recall F1 URL: https://github.com/apache/incubator-mxnet/issues/9589#issuecomment-409030763 @szha Can we remove the `Application` tag, add the `Feature Request` tag instead? Also not sure why the `Autograd` tag is here, are there particular concern regarding autograd with this change? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] piyushghai commented on a change in pull request #11626: [MXNET-651] MXNet Model Backwards Compatibility Checker
piyushghai commented on a change in pull request #11626: [MXNET-651] MXNet
Model Backwards Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#discussion_r206344664
##
File path:
tests/nightly/model_backwards_compatibility_check/train_mxnet_legacy_models.sh
##
@@ -0,0 +1,71 @@
+#!/usr/bin/env bash
+
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+#Author: Piyush Ghai
+
+set -ex
+
+run_models() {
+ echo '=='
+ echo "Running training files and preparing models"
+ echo '=='
+ python model_backwards_compat_train.py
+ echo '=='
+}
+
+install_mxnet() {
+ version=$1
+ echo "Installing MXNet "$version
+ pip install mxnet==$version --user
+}
+
+echo `pwd`
+cd tests/nightly/model_backwards_compatibility_check
+echo `pwd`
+
+## Fetch the latest release tags, filtering out 'rcs' and filtering out some
other irrelevant ones
+## This list is sorted in descending order chronologically.
+## Sample output for the below git tag command is : 1.2.0 utils 1.1.0 1.0.0
0.12.1
+## so from this sample we will pick up the top two : 1.2.0 and 1.1.0 and train
models on them
+## Now while performing inference the latest version could be 1.3.0, which
will help in validating models trained
+## on 1.1.0 and 1.2.0 by loading them on the latest version (1.3.0)
+## Over a period of time, the model repository will grow since with every new
release we
+## upload models trained on newer versions as well through this script
+previous_versions=($(git tag --sort=-creatordate | grep --invert-match rc))
+count=0
+for version in ${previous_versions[*]}
+do
+ # We just need to train the previous two versions. This logic can be
changed later on as welll.
+ if [[ "$count" -gt 1 ]]
+ then
+ echo "Successfully trained files for the previous two MXNet
release versions"
+ exit 0
+ fi
+
+ ## If MXNet major version starts with a number >=1. with a wildcard
match for the minor version numbers
+ ## Could have used a [[:digit:]]+. as well but it was not working as a
traditional regex in bash.
+ ## so had to resort to using [[:digit:]] [[:digit:]]* to indicate
multi-digit version regex match
+ if [[ $version = [[:digit:][[:digit:]]*.[[:digit:]].[[:digit:]] ]]
Review comment:
Done. d8fa75d72bd87466112f58865cebb38883b63171.
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] piyushghai commented on a change in pull request #11626: [MXNET-651] MXNet Model Backwards Compatibility Checker
piyushghai commented on a change in pull request #11626: [MXNET-651] MXNet
Model Backwards Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#discussion_r206344638
##
File path:
tests/nightly/model_backwards_compatibility_check/train_mxnet_legacy_models.sh
##
@@ -0,0 +1,69 @@
+#!/usr/bin/env bash
+
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+#Author: Piyush Ghai
+
+run_models() {
+ echo '=='
+ echo "Running training files and preparing models"
+ echo '=='
+ python model_backwards_compat_train.py
+ echo '=='
+}
+
+install_mxnet() {
+ version=$1
+ echo "Installing MXNet "$version
+ pip install mxnet==$version
+}
+
+install_boto3(){
+ echo "Intalling boto3"
+ pip install boto3
+}
+
+echo `pwd`
+cd tests/nightly/model_backwards_compatibility_check
+echo `pwd`
+
+install_boto3
+
+## Fetch the latest release tags, filtering out 'rcs' and filtering out some
other irrelevant ones
+## This list is sorted in descending order chronologically. Keeping n = 5 for
a precautionary check.
+## Sample output for the below git tag command is : 1.2.0 utils 1.1.0 1.0.0
0.12.1
+previous_versions=($(git tag --sort=-creatordate | grep --invert-match rc |
head -n 5))
+count=0
+for version in ${previous_versions[*]}
+do
+ # We just need to train the previous two versions. This logic can be
changed later on as welll.
+ if [[ "$count" -gt 1 ]]
Review comment:
Done. d8fa75d72bd87466112f58865cebb38883b63171.
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] sad- commented on issue #9064: How to perform 2D convolution with user defined kernel just like sobel filter?
sad- commented on issue #9064: How to perform 2D convolution with user defined kernel just like sobel filter? URL: https://github.com/apache/incubator-mxnet/issues/9064#issuecomment-409046704 @marcoabreu can you close out this issue since the question has been answered with example by @thomelane This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] yzhliu commented on issue #8221: Question about Custom Iterator?
yzhliu commented on issue #8221: Question about Custom Iterator? URL: https://github.com/apache/incubator-mxnet/issues/8221#issuecomment-409047026 Since we don't receive response from the requester for a long time, I close it for now. @danache Feel free to reopen if you're still facing the problem. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] anirudhacharya opened a new pull request #11936: [MXNET-344] Update ONNX docs with the latest supported ONNX version
anirudhacharya opened a new pull request #11936: [MXNET-344] Update ONNX docs with the latest supported ONNX version URL: https://github.com/apache/incubator-mxnet/pull/11936 ## Description ## The current ONNX version supported by MXNet is 1.2.1. Updating the docs accordingly. ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [ ] The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant [JIRA issue](https://issues.apache.org/jira/projects/MXNET/issues) created (except PRs with tiny changes) - [ ] Changes are complete (i.e. I finished coding on this PR) - [ ] All changes have test coverage: - Unit tests are added for small changes to verify correctness (e.g. adding a new operator) - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore) - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL) - [ ] Code is well-documented: - For user-facing API changes, API doc string has been updated. - For new C++ functions in header files, their functionalities and arguments are documented. - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable - Check the API doc at http://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html - [ ] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change ## Comments ## @Roshrini @zhreshold This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sad- commented on issue #8985: Torch Example Requires mx.th.randn and mx.symbol.TorchModule API
sad- commented on issue #8985: Torch Example Requires mx.th.randn and mx.symbol.TorchModule API URL: https://github.com/apache/incubator-mxnet/issues/8985#issuecomment-409048049 @sandeep-krishnamurthy can we close out this issue. example no longer exists on current master This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] azai91 commented on issue #11804: softrelu activation clipping bug in MKLDNN
azai91 commented on issue #11804: softrelu activation clipping bug in MKLDNN URL: https://github.com/apache/incubator-mxnet/issues/11804#issuecomment-409050570 @ZhennanQin do we have an ETA when this would be fixed? should we fallback if input too large? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] eric-haibin-lin commented on issue #11925: mxnet.ndarray.stack : PLEASE UPDATE DOC
eric-haibin-lin commented on issue #11925: mxnet.ndarray.stack : PLEASE UPDATE DOC URL: https://github.com/apache/incubator-mxnet/issues/11925#issuecomment-409064379 @y12uc231 Hi, thanks for the feedback. Indeed it's not very clear. And even worse, it's a common problem for all operator examples. The problem is that the example is written once, and shown in python/scala/R documentation, so we tried to make it more language agnostic but unfortunately becomes very misleading... This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] wkcn commented on issue #11927: Substitute of torch.Tensor.data_ptr()
wkcn commented on issue #11927: Substitute of torch.Tensor.data_ptr() URL: https://github.com/apache/incubator-mxnet/issues/11927#issuecomment-409067222 Using `MXNDArrayGetData` to get the first element of the NDArray. Notice: You should call `wait_to_read` before reading a NDArray and call `wait_to_write` before writing a NDArray. There is no `wait_to_write` in NDArray, so `mx.nd.NDArray.wait_to_write = lambda self : _LIB.MXNDArrayWaitToWrite(self.handle)` Example: https://github.com/wkcn/MobulaOP/blob/master/mobula_op/glue/mx.py#L28 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #7814: Pooling operator expects kernel size even when 'global_pool' is set to True
ThomasDelteil commented on issue #7814: Pooling operator expects kernel size
even when 'global_pool' is set to True
URL:
https://github.com/apache/incubator-mxnet/issues/7814#issuecomment-409020538
@indhub this does not seem necessary anymore. Please close:
```
pool5 = mx.symbol.Pooling(data=mx.sym.var('data'),
pool_type='avg',
global_pool=True,
name='pool')
mx.test_utils.simple_forward(pool5, data=np.ones((1,3,224,224)))
array( 1.]],
[[ 1.]],
[[ 1., dtype=float32)
```
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] ThomasDelteil commented on issue #8636: What differences between ImageRecordIter and ImageRecordIter_v1?
ThomasDelteil commented on issue #8636: What differences between ImageRecordIter and ImageRecordIter_v1? URL: https://github.com/apache/incubator-mxnet/issues/8636#issuecomment-409023173 @sandeep-krishnamurthy @indhub can you please close the issue? Thanks! This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #9588: metric should have TP FP TN FN Precision Recall F1 for both macro and micro versions
ThomasDelteil commented on issue #9588: metric should have TP FP TN FN Precision Recall F1 for both macro and micro versions URL: https://github.com/apache/incubator-mxnet/issues/9588#issuecomment-409030047 @szha looks like #9777 address this ? If so please consider closing. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] piyushghai commented on a change in pull request #11626: [MXNET-651] MXNet Model Backwards Compatibility Checker
piyushghai commented on a change in pull request #11626: [MXNET-651] MXNet
Model Backwards Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#discussion_r206344952
##
File path:
tests/nightly/model_backwards_compatibility_check/train_mxnet_legacy_models.sh
##
@@ -0,0 +1,69 @@
+#!/usr/bin/env bash
+
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+#Author: Piyush Ghai
+
+run_models() {
+ echo '=='
+ echo "Running training files and preparing models"
+ echo '=='
+ python model_backwards_compat_train.py
+ echo '=='
+}
+
+install_mxnet() {
+ version=$1
+ echo "Installing MXNet "$version
+ pip install mxnet==$version
+}
+
+install_boto3(){
+ echo "Intalling boto3"
+ pip install boto3
+}
+
+echo `pwd`
+cd tests/nightly/model_backwards_compatibility_check
+echo `pwd`
+
+install_boto3
+
+## Fetch the latest release tags, filtering out 'rcs' and filtering out some
other irrelevant ones
+## This list is sorted in descending order chronologically. Keeping n = 5 for
a precautionary check.
+## Sample output for the below git tag command is : 1.2.0 utils 1.1.0 1.0.0
0.12.1
+previous_versions=($(git tag --sort=-creatordate | grep --invert-match rc |
head -n 5))
+count=0
+for version in ${previous_versions[*]}
+do
+ # We just need to train the previous two versions. This logic can be
changed later on as welll.
+ if [[ "$count" -gt 1 ]]
+ then
+ echo "Successfully trained files for the previous two MXNet
release versions"
+ exit 1
+ fi
+
+ ## If MXNet major version starts with a number >=1. with a wildcard
match for the minor version numbers
Review comment:
It handles here : d8fa75d72bd87466112f58865cebb38883b63171.
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] sandeep-krishnamurthy commented on issue #11890: Error During Make - undefined reference
sandeep-krishnamurthy commented on issue #11890: Error During Make - undefined reference URL: https://github.com/apache/incubator-mxnet/issues/11890#issuecomment-409035250 Yes it looks like your OpenCV installation and for MXNet to find it is having issue. Can you try commands here - https://github.com/apache/incubator-mxnet/blob/master/setup-utils/install-mxnet-amz-linux.sh#L48 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] gigasquid commented on issue #11808: remove mod from arity 2 version of load-checkpoint in clojure-package
gigasquid commented on issue #11808: remove mod from arity 2 version of load-checkpoint in clojure-package URL: https://github.com/apache/incubator-mxnet/pull/11808#issuecomment-409036971 @jimdunn Thanks so much for finding and fixing this We've been trying to improve the coverage of our unit tests, would you mind adding a test for this arity to make sure it doesn't break again? https://github.com/apache/incubator-mxnet/blob/master/contrib/clojure-package/test/org/apache/clojure_mxnet/module_test.clj#L105 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] safrooze commented on issue #11906: support 1D and 3D arrays in MKLDNN.
safrooze commented on issue #11906: support 1D and 3D arrays in MKLDNN. URL: https://github.com/apache/incubator-mxnet/issues/11906#issuecomment-409037264 @TaoLv By 3D tensor, I meant something with (batch_size, channels, width), which is a 1D convolution. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] Roshrini commented on issue #11932: Not Found Links in the doc of MXNet : Training with multiple GPUs
Roshrini commented on issue #11932: Not Found Links in the doc of MXNet : Training with multiple GPUs URL: https://github.com/apache/incubator-mxnet/issues/11932#issuecomment-409042364 @sandeep-krishnamurthy Could you please add label: Doc, Python This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] Ishitori commented on issue #11144: Save/Load Model?
Ishitori commented on issue #11144: Save/Load Model? URL: https://github.com/apache/incubator-mxnet/issues/11144#issuecomment-409042311 @anchen1011 Did it work for you? If yes, please, close the issue. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] aaronmarkham commented on issue #11921: [MXNET-711] Added updated logos to the powered by page
aaronmarkham commented on issue #11921: [MXNET-711] Added updated logos to the powered by page URL: https://github.com/apache/incubator-mxnet/pull/11921#issuecomment-409044406 @gigasquid This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] lanking520 closed pull request #11621: [MXNET-531] NeuralStyle Example for Scala
lanking520 closed pull request #11621: [MXNET-531] NeuralStyle Example for Scala
URL: https://github.com/apache/incubator-mxnet/pull/11621
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull request (from a fork), the diff is supplied
below (as it won't show otherwise due to GitHub magic):
diff --git
a/scala-package/examples/src/main/scala/org/apache/mxnetexamples/neuralstyle/ModelVgg19.scala
b/scala-package/examples/src/main/scala/org/apache/mxnetexamples/neuralstyle/ModelVgg19.scala
index 4d9aa35d21f..ca4c242ab1c 100644
---
a/scala-package/examples/src/main/scala/org/apache/mxnetexamples/neuralstyle/ModelVgg19.scala
+++
b/scala-package/examples/src/main/scala/org/apache/mxnetexamples/neuralstyle/ModelVgg19.scala
@@ -17,92 +17,73 @@
package org.apache.mxnetexamples.neuralstyle
-import org.apache.mxnet.Context
-import org.apache.mxnet.Executor
-import org.apache.mxnet.NDArray
-import org.apache.mxnet.Symbol
-import org.apache.mxnet.Shape
+import org.apache.mxnet.{Context, Executor, NDArray, Shape, Symbol}
/**
- * Definition for the neuralstyle network and initialize it with pretrained
weight
- * @author Depeng Liang
- */
+ * Definition for the neuralstyle network and initialize it with pretrained
weight
+ */
object ModelVgg19 {
case class ConvExecutor(executor: Executor, data: NDArray, dataGrad: NDArray,
- style: Array[NDArray], content: NDArray, argDict:
Map[String, NDArray])
+ style: Array[NDArray], content: NDArray, argDict:
Map[String, NDArray])
+
+ def ConvRelu(data : Symbol, convName : String, reluName : String,
+ numFilter : Int, kernel : (Int, Int) = (3, 3),
+ stride : (Int, Int) = (1, 1)) : Symbol = {
+val conv = Symbol.api.Convolution(data = Some(data), num_filter =
numFilter,
+ pad = Some(Shape(1, 1)), kernel = Shape(kernel._1, kernel._2),
+ stride = Some(Shape(stride._1, stride._2)), no_bias = Some(false),
+ workspace = Some(1024), name = convName)
+val relu = Symbol.api.relu(data = Some(conv), name = reluName)
+conv.dispose()
+relu
+ }
def getSymbol: (Symbol, Symbol) = {
+getVggSymbol()
+ }
+
+ def getVggSymbol(prefix: String = "", contentOnly: Boolean = false):
(Symbol, Symbol) = {
// declare symbol
-val data = Symbol.Variable("data")
-val conv1_1 = Symbol.Convolution("conv1_1")()(Map("data" -> data ,
"num_filter" -> 64,
-"pad" -> "(1,1)", "kernel" -> "(3,3)",
"stride" -> "(1,1)",
-"no_bias" -> false, "workspace" ->
1024))
-val relu1_1 = Symbol.Activation("relu1_1")()(Map("data" -> conv1_1 ,
"act_type" -> "relu"))
-val conv1_2 = Symbol.Convolution("conv1_2")()(Map("data" -> relu1_1 ,
"num_filter" -> 64,
-"pad" -> "(1,1)", "kernel" -> "(3,3)",
"stride" -> "(1,1)",
-"no_bias" -> false, "workspace" ->
1024))
-val relu1_2 = Symbol.Activation("relu1_2")()(Map("data" -> conv1_2 ,
"act_type" -> "relu"))
-val pool1 = Symbol.Pooling("pool1")()(Map("data" -> relu1_2 , "pad" ->
"(0,0)",
-"kernel" -> "(2,2)", "stride" -> "(2,2)",
"pool_type" -> "avg"))
-val conv2_1 = Symbol.Convolution("conv2_1")()(Map("data" -> pool1 ,
"num_filter" -> 128,
-"pad" -> "(1,1)", "kernel" -> "(3,3)",
"stride" -> "(1,1)",
-"no_bias" -> false, "workspace" ->
1024))
-val relu2_1 = Symbol.Activation("relu2_1")()(Map("data" -> conv2_1 ,
"act_type" -> "relu"))
-val conv2_2 = Symbol.Convolution("conv2_2")()(Map("data" -> relu2_1 ,
"num_filter" -> 128,
-"pad" -> "(1,1)", "kernel" -> "(3,3)",
"stride" -> "(1,1)",
-"no_bias" -> false, "workspace" ->
1024))
-val relu2_2 = Symbol.Activation("relu2_2")()(Map("data" -> conv2_2 ,
"act_type" -> "relu"))
-val pool2 = Symbol.Pooling("pool2")()(Map("data" -> relu2_2 , "pad" ->
"(0,0)",
-"kernel" -> "(2,2)", "stride" -> "(2,2)",
"pool_type" -> "avg"))
-val conv3_1 = Symbol.Convolution("conv3_1")()(Map("data" -> pool2 ,
"num_filter" -> 256,
-"pad" -> "(1,1)", "kernel" -> "(3,3)",
"stride" -> "(1,1)",
-"no_bias" -> false, "workspace" ->
1024))
-val relu3_1 = Symbol.Activation("relu3_1")()(Map("data" -> conv3_1 ,
"act_type" -> "relu"))
-val conv3_2 = Symbol.Convolution("conv3_2")()(Map("data" -> relu3_1 ,
"num_filter" -> 256,
-"pad" -> "(1,1)", "kernel" -> "(3,3)",
"stride" -> "(1,1)",
-"no_bias" -> false,
[GitHub] lanking520 opened a new pull request #11621: [MXNET-531] NeuralStyle Example for Scala
lanking520 opened a new pull request #11621: [MXNET-531] NeuralStyle Example for Scala URL: https://github.com/apache/incubator-mxnet/pull/11621 ## Description ## This is the example written in new API in Scala. It include the following three tests: - boost inference - boost training - neural style @yzhliu @nswamy @andrewfayres ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [x] The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant [JIRA issue](https://issues.apache.org/jira/projects/MXNET/issues) created (except PRs with tiny changes) - [x] Add `README` - [x] Changes are complete (i.e. I finished coding on this PR) - [x] All changes have test coverage: - [x] Code is well-documented: - [x] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] lanking520 closed pull request #11844: [MXNET-689] add DataDesc type for the Scala Package
lanking520 closed pull request #11844: [MXNET-689] add DataDesc type for the
Scala Package
URL: https://github.com/apache/incubator-mxnet/pull/11844
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull request (from a fork), the diff is supplied
below (as it won't show otherwise due to GitHub magic):
diff --git a/scala-package/core/src/main/scala/org/apache/mxnet/DType.scala
b/scala-package/core/src/main/scala/org/apache/mxnet/DType.scala
index 4458a7c7aeb..b015bd2169b 100644
--- a/scala-package/core/src/main/scala/org/apache/mxnet/DType.scala
+++ b/scala-package/core/src/main/scala/org/apache/mxnet/DType.scala
@@ -35,4 +35,13 @@ object DType extends Enumeration {
case DType.Unknown => 0
}
}
+ private[mxnet] def getType(dtypeStr: String): DType = {
+dtypeStr match {
+ case "UInt8" => DType.UInt8
+ case "Int32" => DType.Int32
+ case "Float16" => DType.Float16
+ case "Float32" => DType.Float32
+ case "Float64" => DType.Float64
+}
+ }
}
diff --git a/scala-package/core/src/main/scala/org/apache/mxnet/IO.scala
b/scala-package/core/src/main/scala/org/apache/mxnet/IO.scala
index 47fd4eee939..928606db8b7 100644
--- a/scala-package/core/src/main/scala/org/apache/mxnet/IO.scala
+++ b/scala-package/core/src/main/scala/org/apache/mxnet/IO.scala
@@ -19,13 +19,13 @@ package org.apache.mxnet
import org.apache.mxnet.Base._
import org.apache.mxnet.DType.DType
+import org.apache.mxnet.Layout.Layout
import org.apache.mxnet.io.{MXDataIter, MXDataPack}
import org.slf4j.LoggerFactory
import scala.annotation.varargs
import scala.collection.immutable.ListMap
import scala.collection.mutable.ListBuffer
-
/**
* IO iterators for loading training & validation data
*/
@@ -106,7 +106,15 @@ object IO {
checkCall(_LIB.mxDataIterCreateIter(handle, keys, vals, out))
val dataName = params.getOrElse("data_name", "data")
val labelName = params.getOrElse("label_name", "label")
-new MXDataIter(out.value, dataName, labelName)
+val dataLayout = params.getOrElse("dataLayout", "NCHW")
+val labelLayout = params.getOrElse("labelLayout", "N")
+val dataDType = params.getOrElse("dataDType", "Float32")
+val labelDType = params.getOrElse("labelDType", "Int32")
+new MXDataIter(out.value, dataName, labelName,
+ dataLayout = Layout.getLayout(dataLayout),
+ labelLayout = Layout.getLayout(labelLayout),
+ dataDType = DType.getType(dataDType),
+ labelDType = DType.getType(labelDType))
}
// Convert data into canonical form.
@@ -140,7 +148,11 @@ class DataBatch(val data: IndexedSeq[NDArray],
// use ListMap to indicate the order of data/label loading
// (must match the order of input data/label)
private val providedData: ListMap[String, Shape] = null,
-private val providedLabel: ListMap[String, Shape] = null) {
+private val providedLabel: ListMap[String, Shape] = null,
+val dataDType: DType = Base.MX_REAL_TYPE,
+val labelDType: DType = DType.Int32,
+val dataLayout: Layout = Layout.NCHW,
+val labelLayout: Layout = Layout.N) {
/**
* Dispose its data and labels
* The object shall never be used after it is disposed.
@@ -170,6 +182,10 @@ object DataBatch {
private var label: IndexedSeq[NDArray] = null
private var index: IndexedSeq[Long] = null
private var pad: Int = 0
+private var dataLayout: Layout = Layout.NCHW
+private var labelLayout: Layout = Layout.N
+private var dataDType: DType = Base.MX_REAL_TYPE
+private var labelDType: DType = DType.Int32
private var bucketKey: AnyRef = null
private var datatShapes: ListMap[String, Shape] = null
private var labelShapes: ListMap[String, Shape] = null
@@ -216,6 +232,30 @@ object DataBatch {
this
}
+/**
+ * Set the dtype.
+ * @param dataDType The dtype of the data, default is Float32
+ * @param labelDType The dtype of the label, default is Int32
+ * @return this
+ */
+def setDType(dataDType: DType, labelDType: DType): Builder = {
+ this.dataDType = dataDType
+ this.labelDType = labelDType
+ this
+}
+
+/**
+ * Set the layout.
+ * @param dataLayout The layout of the data, default is NCHW
+ * @param labelLayout The layout of the label, default is N
+ * @return this
+ */
+def setLayout(dataLayout: Layout, labelLayout: Layout): Builder = {
+ this.dataLayout = dataLayout
+ this.labelLayout = labelLayout
+ this
+}
+
/**
* Set the bucket key, used for bucketing module.
* @param bucketKey the bucket key related to this batch.
@@ -258,7 +298,8 @@ object DataBatch {
def build(): DataBatch = {
require(data != null, "data is required.")
-
[GitHub] ThomasDelteil commented on issue #9959: Revise the profiler examples and documents
ThomasDelteil commented on issue #9959: Revise the profiler examples and documents URL: https://github.com/apache/incubator-mxnet/issues/9959#issuecomment-409025832 @sxjscience The examples have been updated in https://github.com/apache/incubator-mxnet/pull/10190 The performance has been updated in https://github.com/apache/incubator-mxnet/pull/10761 @sxjscience could you please close the issue? Thanks This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sandeep-krishnamurthy closed pull request #11910: Improving documentation and error messages for Async distributed training with Gluon
sandeep-krishnamurthy closed pull request #11910: Improving documentation and
error messages for Async distributed training with Gluon
URL: https://github.com/apache/incubator-mxnet/pull/11910
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull request (from a fork), the diff is supplied
below (as it won't show otherwise due to GitHub magic):
diff --git a/docs/faq/distributed_training.md b/docs/faq/distributed_training.md
index 70078ba6095..d4fa72db23a 100644
--- a/docs/faq/distributed_training.md
+++ b/docs/faq/distributed_training.md
@@ -73,6 +73,23 @@ These can be passed as arguments to the iterator.
You can look at
[example/gluon/image_classification.py](https://github.com/apache/incubator-mxnet/blob/master/example/gluon/image_classification.py)
to see an example usage.
+### Updating weights
+KVStore server supports two modes, one which aggregates the gradients and
updates the weights using those gradients, and second where the server only
aggregates gradients. In the latter case, when a worker process pulls from
kvstore, it gets the aggregated gradients. The worker then uses these gradients
and applies the weights locally.
+
+When using Gluon there is an option to choose between these modes by passing
`update_on_kvstore` variable when you create the
[Trainer](https://mxnet.incubator.apache.org/versions/master/api/python/gluon/gluon.html#mxnet.gluon.Trainer)
object like this:
+
+```
+trainer = gluon.Trainer(net.collect_params(), optimizer='sgd',
+optimizer_params={'learning_rate': opt.lr,
+ 'wd': opt.wd,
+ 'momentum': opt.momentum,
+ 'multi_precision': True},
+kvstore=kv,
+update_on_kvstore=True)
+```
+
+When using the symbolic interface, it performs the weight updates on the
server without the user having to do anything special.
+
### Different Modes of Distributed Training
Distributed training itself is enabled when kvstore creation string contains
the word `dist`.
@@ -86,9 +103,9 @@ In this mode, if a worker crashes, then it halts the
progress of all workers.
- `dist_async`: In asynchronous distributed training, the server receives
gradients from one worker and immediately updates its store, which it uses to
respond to any future pulls.
This means that a worker who finishes processing a batch can pull the current
parameters from server and start the next batch,
even if other workers haven't finished processing the earlier batch.
-This is faster than `dist_sync` but can take more epochs to converge.
-In `async` mode, it is required to pass an optimizer because in the absence of
an optimizer kvstore would replace the stored weights with received weights and
this doesn't make sense for training in asynchronous mode.
+This is faster than `dist_sync` because there is no cost of synchronization,
but can take more epochs to converge.
The update of weights is atomic, meaning no two updates happen on the same
weight at the same time. However, the order of updates is not guaranteed.
+In `async` mode, it is required to pass an optimizer because in the absence of
an optimizer kvstore would replace the stored weights with received weights and
this doesn't make sense for training in asynchronous mode. Hence, when using
Gluon with `async` mode we need to set `update_on_kvstore` to `True`.
- `dist_sync_device`: Same as `dist_sync` except that when there are multiple
GPUs being used on each node,
this mode aggregates gradients and updates weights on GPU while dist_sync does
so on CPU memory.
diff --git a/python/mxnet/gluon/trainer.py b/python/mxnet/gluon/trainer.py
index b4263410a50..98a6878b94b 100644
--- a/python/mxnet/gluon/trainer.py
+++ b/python/mxnet/gluon/trainer.py
@@ -187,6 +187,11 @@ def _init_kvstore(self):
arg_arrays = {param.name: param.data(self._contexts[0]) for param
in self._params}
kvstore, update_on_kvstore = _create_kvstore(config['kvstore'],
len(self._contexts),
arg_arrays)
+if kvstore and 'async' in kvstore.type and
config['update_on_kvstore'] is not None\
+and not config['update_on_kvstore']:
+raise ValueError("Please set update_on_kvstore to true "
+ "when training in async mode.")
+
if config['update_on_kvstore'] is not None:
update_on_kvstore = config['update_on_kvstore']
if kvstore:
@@ -195,7 +200,8 @@ def _init_kvstore(self):
self._distributed = 'dist' in kvstore.type
if self._distributed:
# kv.pull(row_sparse_grad) is not supported for dist kvstore
-update_on_kvstore =
[GitHub] aaronmarkham commented on issue #11438: docs using xml in code blocks will improperly render
aaronmarkham commented on issue #11438: docs using xml in code blocks will improperly render URL: https://github.com/apache/incubator-mxnet/issues/11438#issuecomment-409032714 Actually there is. Something keeps mucking with the code and escaping it improperly. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ThomasDelteil commented on issue #5115: when i run the train_alternate.py cv2.error , but i test opencv indenpendenly it runs well
ThomasDelteil commented on issue #5115: when i run the train_alternate.py cv2.error , but i test opencv indenpendenly it runs well URL: https://github.com/apache/incubator-mxnet/issues/5115#issuecomment-409032755 @indhub can you please close this issue? Thanks! This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sandeep-krishnamurthy closed issue #9157: YARN support data locality
sandeep-krishnamurthy closed issue #9157: YARN support data locality URL: https://github.com/apache/incubator-mxnet/issues/9157 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] piyushghai commented on a change in pull request #11626: [MXNET-651] MXNet Model Backwards Compatibility Checker
piyushghai commented on a change in pull request #11626: [MXNET-651] MXNet
Model Backwards Compatibility Checker
URL: https://github.com/apache/incubator-mxnet/pull/11626#discussion_r206347507
##
File path:
tests/nightly/model_backwards_compatibility_check/train_mxnet_legacy_models.sh
##
@@ -0,0 +1,82 @@
+#!/usr/bin/env bash
+
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+#Author: Piyush Ghai
+
+set -ex
+
+run_models() {
+ echo '=='
+ echo "Running training files and preparing models"
+ echo '=='
+ python model_backwards_compat_train.py
+ echo '=='
+}
+
+install_mxnet() {
+ version=$1
+ echo "Installing MXNet "$version
+ pip install mxnet==$version --user
+}
+
+## Cuts the string and gives only the major version part.
+## eg : 12.3.0 ---> 12
+get_major_version() {
+major=$(echo $1 | cut -d. -f1)
+echo $major
+}
+
+## We read the current major version from libinfo.py file. And we extract the
major version from it.
+curr_mxnet_version=$(grep -w "__version__" python/mxnet/libinfo.py | grep -o
'".*"' | sed 's/"//g')
+curr_major_version=$(get_major_version $curr_mxnet_version)
Review comment:
Done. ca01aa2bd9a183af10a4c3ae7ca5194c395b4ffb.
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] sad- commented on issue #4709: initialization error when finetune a pretrained cnn with lstm-ctc
sad- commented on issue #4709: initialization error when finetune a pretrained cnn with lstm-ctc URL: https://github.com/apache/incubator-mxnet/issues/4709#issuecomment-409037873 @xinghedyc could you close this issue if your original question has been answered? thanks This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] Ishitori commented on issue #7958: Engine shutdown
Ishitori commented on issue #7958: Engine shutdown URL: https://github.com/apache/incubator-mxnet/issues/7958#issuecomment-409037696 @Paul0M can you provide a minimal reproducible example with how to run it? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] yzhliu closed issue #9935: Matplotlib plot crashes Jupyter Notebook with segmentation fault: 11
yzhliu closed issue #9935: Matplotlib plot crashes Jupyter Notebook with segmentation fault: 11 URL: https://github.com/apache/incubator-mxnet/issues/9935 This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sad- commented on issue #9911: The different between rescale_grad on optimizer and normalization in loss layer.
sad- commented on issue #9911: The different between rescale_grad on optimizer and normalization in loss layer. URL: https://github.com/apache/incubator-mxnet/issues/9911#issuecomment-409046382 @huyangc could you close the issue if you question is answered? or ask a specific follow up question if you have one. thanks This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] yzhliu commented on issue #9935: Matplotlib plot crashes Jupyter Notebook with segmentation fault: 11
yzhliu commented on issue #9935: Matplotlib plot crashes Jupyter Notebook with segmentation fault: 11 URL: https://github.com/apache/incubator-mxnet/issues/9935#issuecomment-409046086 @Ishitori Thanks for checking. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] piiswrong commented on issue #11899: non-deterministic backward of scatter_nd
piiswrong commented on issue #11899: non-deterministic backward of scatter_nd URL: https://github.com/apache/incubator-mxnet/issues/11899#issuecomment-409048465 Why do we need to this to be deterministic? Is there a use case? Note that we don't promise determinism for most of the operators. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] piiswrong edited a comment on issue #11899: non-deterministic backward of scatter_nd
piiswrong edited a comment on issue #11899: non-deterministic backward of scatter_nd URL: https://github.com/apache/incubator-mxnet/issues/11899#issuecomment-409048465 Why do we need to this to be deterministic? Is there a use case? Note that we don't promise determinism by default. This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] sad- commented on issue #10237: No setupenv.cmd present in the folder
sad- commented on issue #10237: No setupenv.cmd present in the folder URL: https://github.com/apache/incubator-mxnet/issues/10237#issuecomment-409050071 truly there is no setupenv.cmd present in the folder. see https://discuss.mxnet.io/t/installing-mxnet-on-windows/374 if you're still having issues installing or trying installing from source. the installation docs need to be updated to correct the stale info This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] zhreshold commented on issue #11908: Fix shared memory with gluon dataloader, add option pin_memory
zhreshold commented on issue #11908: Fix shared memory with gluon dataloader, add option pin_memory URL: https://github.com/apache/incubator-mxnet/pull/11908#issuecomment-409051889 @piiswrong Can you have a look? This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
